CoChat
Sign Up
Cited by 391Language Development and DisordersReading and Literacy DevelopmentEarly Childhood Education and Development

Shared Picture Book Reading Interventions for Child Language Development: A Systematic Review and Meta-Analysis

Open access
Dowdall, Nicholas, Melendez-Torres, G J, Murray, Lynne, Gardner, Frances, Hartford, Leila, Cooper, Peter J·2020·Child Development
DOI 10.1111/cdev.13225
Open this paper in CoChat — free accountOpen PDF

Abstract

Interventions that train parents to share picture books with children are seen as a strategy for supporting child language development. We conducted meta-analyses using robust variance estimation modeling on results from 19 RCTs (Ntotal = 2,594; Mchildage = 1–6 years). Overall, book-sharing interventions had a small sized effect on both expressive language (d = 0.41) and receptive language (d = 0.26). They had a large effect on caregiver book-sharing competence (d = 1.01).

Integrity signals

Registry

Crossref · 10.1111/cdev.13225

Cites

63 of 63 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

14 of 77 cited sources checked · 5 supported · 0 flagged · accuracy not yet measured

  • Premisecitation

    Two studies in an impoverished urban South African township provided strong evidence for the efficacy of dialogic book-sharing in low- and middle-income settings.

    “Most recently, two studies conducted in an impoverished urban South African township provided strong evidence for the efficacy of dialogic book-sharing in LMIC settings (Cooper et al., 2014;Vally, Murray, Tomlinson, & Cooper, 2015).”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Cooper et al., 2014 source
    • cites Vally, Murray, Tomlinson, & Cooper, 2015 source
  • Premisecitation

    In socio-economically disadvantaged contexts, caregivers engage in substantially less shared reading with their children.

    “in such contexts caregivers engage in substantially less shared reading with their children (Berkule, Dreyer, Huberman, Fierman, & Mendelsohn, 2007;Cooper et al., 2014;Le Roux, 2012).”
    Quote in the paperSubstantial claim2 of 3 sources checked
    • cites Berkule et al., 2007 source

      The Discussion (page unavailable) reports prior studies associating reduced income and lower maternal education with less frequent shared reading. But this paper’s own study measured mothers’ plans and resources, not how often they read with their children, and the cited discussion does not establish that the difference was substantial. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
    • cites Cooper et al., 2014 source
    • cites Le Roux, 2012

      Section 2.3.2 links financial difficulties and lack of books to parents’ limited involvement in children’s reading. The passage discussing Tiemensma (2008:62) also says inadequate social and economic support puts children at risk of never sharing a book with a parent. Neither passage establishes the comparative claim that caregivers in disadvantaged contexts engage in substantially less shared reading. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
  • Premisecitation

    Families facing high adversity are less likely than more advantaged families to use a book-sharing style associated with positive language outcomes.

    “families experiencing high levels of adversity are less likely than more advantaged families to use a style of book-sharing associated with positive language outcomes (Berkule et al., 2007;Bornstein & Putnick, 2012;El Moussaoui & Braster, 2011;Van Kleeck, 1998).”
    Quote in the paperSubstantial claim3 of 3 sources checked
    • cites Berkule et al., 2007 source

      The Discussion (page not provided) says mothers with lower education or socioeconomic status were less likely to report plans to share books in infancy or to have baby books at home. But the study assessed plans and resources among mothers of newborns, not whether families used a particular book-sharing style or whether that style produced positive language outcomes. (LLM, conf=0.87)

      Experimental — accuracy not yet measured
    • cites Bornstein & Putnick, 2012 source

      The Results (Cognitive Caregiving: Deviation from the Grand Mean; page unavailable) report that mothers in low-HDI countries were less likely to read to their children than those in high-HDI countries. The Discussion (Implications for Policy; page unavailable) links more frequent reading with higher reported vocabulary. The paper measures whether reading occurred, not the style of book-sharing, so it does not establish that families facing adversity use a particular style less often. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites El Moussaoui & Braster, 2011 source

      In “Reading as a Stimulating Activity Relevant to Cognitive Development” (page unavailable), the paper reports that all mothers who could not read and four less-educated mothers never read to their children, while discussing reading among more-educated mothers. This supports a difference in reading practices by education, but the paper does not establish a difference in book-sharing style by adversity level or link a particular style to positive language outcomes. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
  • Premisecitation

    At least one formal mediation analysis showed that improvements in caregiver book-sharing competence associated with training accounted for much of the improvement in child language.

    “There has been at least one formal mediation analysis (Murray et al., 2016) that showed how improvements in caregiver book-sharing competence associated with training accounted for much of the improvement in child language.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Murray et al., 2016 source
  • Study design and conductreasoning

    In studies relying on parent-reported child language, blinding mattered because awareness of intervention status could skew answers.

    “However, for studies that used parent self-report measures of child language (Cronan et al., 1996;Vally et al., 2015), blinding was a potentially important issue because knowledge of having received the intervention could have biased responses.”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Cronan et al., 1996 source
    • cites Vally et al., 2015 source

      The Child language section says the study used mothers’ reports to measure child language. The Discussion raises a potential reporting problem: trained carers may have become more aware of their children’s linguistic abilities, rather than the children’s language having improved. It does not specifically attribute possible bias to carers’ knowledge of their intervention status. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
  • Classified as probably not a substantial claim. 49
  • May combine two claims. 48
  • The quote only partly states this claim. 1

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    Publication bias was assessed by plotting each trial's effect against standard errors, creating one effect size per study per outcome.

    “To assess risk of publication bias, we plotted each trial's effect against the standard errors (Sterne et al., 2011), creating a single effect size per study per outcome.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Sterne et al., 2011 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Early developmental and language deficits are concentrated in low- and middle-income countries, especially Sub-Saharan Africa and Asia.

    “This is a particularly important question, given that the burden of early developmental and language deficit falls in LMICpredominantly in Sub-Saharan Africa and Asia (Black et al., 2017).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Black et al., 2017 source

      The section “Estimates of children at risk of not attaining developmental potential” (page unavailable) reports a large burden among children under 5 in LMICs, a high prevalence in sub-Saharan Africa, and a large number at risk in South Asia. It measures risk using stunting and extreme poverty, however, rather than establishing the geographic distribution of observed developmental or language deficits; it also specifies South Asia, not Asia generally. (LLM, conf=0.87)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    An estimated more than 250 million children are at risk of not reaching their developmental potential.

    “it is estimated that in excess of 250 million children are at risk of failing to meet their developmental potential (Black et al., 2017).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Black et al., 2017 source

      The section ‘Estimates of children at risk of not attaining developmental potential’ reports that 249.4 million children under 5 in low- and middle-income countries were at risk in 2010. This supports the approximate scale of the claim, but not ‘in excess of 250 million’; the section’s higher estimate of 279.1 million refers to 2004. (LLM, conf=0.88)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    Murray et al. (2016) reported that caregiver sensitivity gains mediated receptive language effects, while caregiver-child reciprocity mediated effects on both expressive and receptive language (and focal attention).

    “Murray and colleagues (2016) found that improvements in caregiver sensitivity during book-sharing significantly mediated the impact of the intervention on receptive language, while improvements in caregiver-child reciprocity during book-sharing mediated the impact of the intervention on both child expressive language and receptive language (and child focal attention).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Murray et al., 2016 source

  • May combine two claims.

    Findingown data

    The effect of book-sharing intervention on child language was moderated by intervention dosage, and lower dosage was linked to minimal impact.

    “The impact of the intervention on child language was moderated by intervention dosage, with lower dosage associated with a minimal impact.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Most studies (k=14) used blind assessors, so outcome assessment was generally well blinded.

    “Outcome assessment was generally well blinded, with the majority of studies (k=14) having blind assessors.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Details on allocation concealment were often insufficient; 10 studies had unclear risk of bias.

    “Allocation concealment was often not reported in sufficient detail: 10 studies were judged as having an unclear risk of bias.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Most interventions (k=15) were delivered in person.

    “The majority of interventions (k=15) were carried out in-person.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    All studies reported on typically developing children only, except Dale (1996) and Towson (2014), who included children with mild to moderate language delays and special education needs respectively.

    “All studies reported on typically developing children only, apart from Dale (1996) and Towson (2014), who included children with mild to moderate language delays and with special education needs respectively.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Dale, 1996 source
    • cites Towson, 2014

  • May combine two claims.

    Findingown data

    In the medium-to-high intensity group, receptive language showed d=0.34 (95% CI 0.17 to 0.50, p=0.002); in the low-intensity group, d=0.09 (95% CI -0.25 to 0.45, p=0.37).

    “Similarly, for those receiving medium to high intensity interventions, the effect size for receptive language was d=0.34 (CI: 0.17,0.50,p=0.002): and for those receiving low intensity interventions it was d=0.09 (CI: -0.25, 0.45; p=0.37).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

    Findingown data

    According to the caregiver education categorization, eight studies had majority tertiary education, 10 did not, and one did not report caregiver education.

    “According to this categorization, eight studies had caregivers with majority tertiary education and 10 studies without this level of caregiver education.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    A single study delivered the intervention through a 20-minute instructional phone call.

    “One study administered the intervention in a 20-minute instructional phone call (Wing-Yin Chow & McBride-Chang, 2003).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Wing-Yin Chow & McBride-Chang, 2003 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    The systematic search yielded 6,448 eligible studies after duplicates were removed.

    “In total, the search yielded 6,448 eligible studies, after duplicates had been removed.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In the medium-to-high intensity group, expressive language showed d=0.54 (95% CI 0.26 to 0.82, p=0.002); in the low-intensity group, d=0.19 (95% CI -0.10 to 0.50, p=0.15).

    “For those receiving medium to high intensity interventions, the effect size for expressive language was d = 0.54 (CI: 0.26, 0.82; p=0.002), whereas for those receiving the low intensity interventions it was d=0.19 (CI:-0.10,0.50; p=0.15).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Just over half of the interventions (10 of 19 studies) involved multiple sessions with training time exceeding 60 minutes with caregivers.

    “Just over half of the interventions (i.e. 10 out of the 19 studies) involved multiple sessions with training time in excess of 60 minutes with caregivers.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    The search identified 6,411 studies via electronic search and 37 by hand searching reference lists of prior reviews and recent evaluation studies.

    “Of these, 6,411 were identified via the electronic search and a further 37 by hand searching the reference lists of the two relevant prior reviews and meta-analyses (Bus et al., 1995;Mol et al., 2008) as well as recent evaluation studies.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Bus et al., 1995 source
    • cites Mol et al., 2008 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Three interventions were administered in video format.

    “Three of the interventions were administered in video format.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Egger's tests were non-significant for expressive language (t=1.73, p=0.11), receptive language (t=1.22, p=0.24), and caregiver competence (t=2.08, p=0.11).

    “This held across all three outcomes: expressive language (t= 1.73; p=0.11), receptive language (t=1.22;p=0.24), and caregiver competence (t=2.08;p=0.11).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Child age at study start ranged from 8.85 months to 66 months.

    “Child age at the start of the studies ranged from 8.85 months (Goldfeld et al., 2012) to 66 months (Sim et al., 2014).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Goldfeld et al., 2012 source
    • cites Sim et al., 2014 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Of included studies, 11 were conducted in the United States; the remaining eight were in other high-income countries and middle-income countries.

    “Of the included studies, 11 were conducted in the United States, while the remaining eight studies were distributed across other high-income countries (including two each in Hong Kong and Australia) and middle-income countries (South Africa (two), Brazil, Turkey).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Moderator analyses with age as a continuous variable produced very similar non-significant results.

    “For clarity, moderator analyses were also run with age as a continuous variable (in years).These produced very similar non-significant results.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Six studies reported caregiver competence outcome data.

    “Outcome data on caregiver competence were reported in six studies.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    No clear moderation of expressive or receptive language effects was found for child age or caregiver education.

    “There was no clear evidence that child age or caregiver education moderated the effect of the intervention on either expressive or receptive language outcomes”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Reporting of socio-demographic variables varied widely, precluding construction of a socioeconomic status variable; the data indicated SES was essentially co-extensive with parental education, which was more reliably reported.

    “Reporting varied widely with regard to socio-demographic variables, which precluded the construction of a socioeconomic status variable (but the data indicated that socio-economic status was essentially co-extensive with parental education, which was more reliably reported).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    There was no evidence that country income level was associated with child receptive language; interventions were just as effective in LMIC as in HIC.

    “There was no evidence for country income level having any association with child receptive language: the interventions were just as effective in LMIC as they were in HIC.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In the 10 longer-duration studies (over 90 minutes), effect sizes were d=0.54 for expressive and d=0.34 for receptive language.

    “When only the 10 studies with longer duration of intervention (i.e., over 90 minutes) were considered, these effect sizes were d=0.54 and d=0.34 respectively.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    Group delivery outperformed individual delivery on both child language outcomes.

    “A potentially important finding was that group-based interventions were more effective than one-on-one interventions for both child language outcomes.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Few studies reported caregiver book-sharing competence outcomes, which precluded a formal subgroup analysis of implementation fidelity.

    “Unfortunately, few studies reported on such outcomes, which precluded a formal subgroup analysis of implementation fidelity.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    More intensive caregiver participation was linked to significantly greater expressive language gains; a similar trend appeared for receptive language.

    “children benefitted significantly more in their expressive language when their caregivers had taken part in a more intensive intervention; and there was a trend for the same association for child receptive language.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    The interventions took diverse forms, producing a degree of clinical heterogeneity.

    “The interventions, while all sharing certain common messages to help parents optimise shared reading practice, came in a variety of forms, which meant that there was a degree of clinical heterogeneity present, reflecting this variability.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    Both clinical and statistical heterogeneity existed among the included studies.

    “clinical and statistical heterogeneity were present within the set of included studies.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Child age did not predict gains in expressive or receptive language; benefit was no different for those younger or older than 3 years.

    “Contrary to an earlier suggestion (Mol et al., 2008), we did not find child age to be related to improvements in either expressive or receptive language: the extent to which language skills benefitted was no different for those younger or older than 3 years.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Mol et al., 2008 source

  • May combine two claims.

    Findingnone given

    Nearly all of those language measures were standardized and well validated.

    “Although almost all of these were standardized and well validated instruments”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Excluding the Goldfeld (2012) study left the overall pattern unchanged, but receptive language became non-significant (p=0.12).

    “The results when this study was excluded did not alter the overall pattern of the findings, however the result for receptive language was no longer statistically significant (p=0.12).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Video-based book-sharing programs have been associated with increased frequency of book sharing and caregiver enjoyment.

    “as well as an increase in the frequency of book sharing and caregiver enjoyment (Golova, Alario, Vivier, Rodriguez, & High, 1999).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Golova et al., 1999 source

  • May combine two claims.

    Premisecitation

    The 'Read together talk together' instructional video, evaluated in a small-scale RCT in Australia, showed positive effects on parental book-sharing techniques and child language.

    “It revealed positive effects on both parental book-sharing techniques and child language (Blom-Hoffman et al., 2006).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Blom-Hoffman et al., 2006 source

  • May combine two claims.

    Premisecitation

    During shared reading, compared with activities such as play or meals, mothers tend to label objects more often and children receive more consistent and informative feedback.

    “during shared reading mothers tend to label objects more frequently and children are provided with more consistent and informative feedback (Snow & Goldfield, 1983;Wells, 1985).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Snow & Goldfield, 1983 source
    • cites Wells, 1985

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim.

    Premisecitation

    Sharing books with young children has been regarded as a particularly effective way to scaffold young children's cognition and language.

    “This has been considered a particularly effective technique for scaffolding a young child's cognition and language (Van Kleeck, 1998)”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Van Kleeck, 1998 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Children's picture books have specific features that make them an especially powerful vehicle for language learning.

    “children's picture books embody specific features that make them an especially powerful vehicle for language learning (Murray, 2014).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Murray, 2014

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Dialogic techniques include asking open-ended questions that go beyond the book content and giving encouragement and praise for child participation.

    “Importantly, what distinguishes these techniques from simply reading to a child is the interactive quality and use of evocative behaviors.This includes asking open-ended questions that extend beyond the content of the book, and providing encouragement and praise for child participation (Lonigan & Whitehurst, 1998;Whitehurst et al., 1988).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Lonigan & Whitehurst, 1998 source
    • cites Whitehurst et al., 1988 source

  • May combine two claims.

    Premisecitation

    In a Mexico kindergarten-based programme for two-year-olds, intervention children showed significantly better expressive and receptive vocabulary and longer, more complex sentences than a no-intervention control group.

    “Compared to a no intervention control group, children in the intervention group demonstrated significantly better expressive and receptive vocabulary following training, as well as the use of longer and more complex sentences.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Valdez-Menchaca & Whitehurst, 1992 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Video formats have been effective in delivering other parenting programs.

    “Video formats have proven effective in the delivery of other parenting programs (Webster-Stratton, 1984).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Webster-Stratton, 1984 source

  • May combine two claims.

    Premisenone given

    Most research and evaluation of book-sharing interventions has occurred in the U.S. and other high-income countries, with very little in low- and middle-income countries.

    “Most research and evaluation of book-sharing interventions has taken place in the U.S. and other high-income countries.There has been very little research on book-sharing conducted in low and middle income countries (LMICs).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Dialogic techniques differ from simply reading to a child by their interactive quality and use of evocative behaviors.

    “Importantly, what distinguishes these techniques from simply reading to a child is the interactive quality and use of evocative behaviors.This includes asking open-ended questions that extend beyond the content of the book, and providing encouragement and praise for child participation (Lonigan & Whitehurst, 1998;Whitehurst et al., 1988).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Lonigan & Whitehurst, 1998 source
    • cites Whitehurst et al., 1988 source

  • May combine two claims.

    Premisecitation

    The registered protocol specified child socio-emotional outcomes as a secondary outcome, but a recent meta-analysis focused directly on that question, making the analysis redundant.

    “While this registered protocol also specified assessment of child socio-emotional outcomes as a secondary outcome, a recent meta-analysis (Xie, Chan, Ji, & Chan, 2018) focussed directly on this question, rendering this analysis redundant.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Xie, Chan, Ji, & Chan, 2018 source

  • May combine two claims.

    Premiseown data

    Bus et al. (1995) included only cross-sectional observational research and did not consider intervention studies.

    “Importantly, however, this review included only cross-sectional observational research and did not consider intervention studies.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Bus et al., 1995 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Mol et al. (2008), a meta-analysis of 16 studies, reported an overall effect of dialogic reading interventions on child vocabulary outcomes of d = 0.42, CI 0.30 to 0.53.

    “This review reported an overall effect of dialogic reading interventions on child vocabulary outcomes of d=0.42,CI: 0.30, 0.53.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mol et al., 2008 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Two reviewers extracted data independently, and discrepancies were resolved through discussion with a third reviewer.

    “Two reviewers (ND and LH) extracted data independently into excel and discrepancies were identified and resolved through discussion with a third reviewer (GM).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    Egger's regression asymmetry tests were used to quantify visible bias.

    “We used Egger's regression asymmetry tests to quantify any visible bias (Egger, Smith, Schneider, & Minder, 1997).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Egger, Smith, Schneider, & Minder, 1997 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Reference lists of known RCTs and previous relevant reviews were scanned for further studies.

    “Reference lists of known RCTs, as well as previous relevant reviews, were scanned for further studies that may have been missed in the initial electronic search.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    For cluster-randomized trials, sample size was reduced to effective sample size using the intra-cluster correlation coefficient to account for the design effect.

    “To account for the design effect, the sample size of cluster-randomized trials was reduced to the effective sample size by using the intra-cluster correlation coefficient.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Mol et al. (2008) found that younger preschool children (2-3 years) benefited more from parental interventions than older kindergarten children (4-5 years), with d = 0.50 and 0.14 respectively.

    “It also showed, as had the review of observational studies by Bus and colleagues (1995), the importance of child age, with younger, preschool children (2-3 year olds) benefitting more from these parental interventions than older children in kindergarten (4-5 year olds) (i.e.d= 0.50 and 0.14 respectively).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mol et al., 2008 source

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductnone given

    Articles were independently screened by two researchers and sequentially excluded according to eligibility criteria.

    “Articles were independently screened by two researchers (ND and LH) and sequentially excluded according to the eligibility criteria.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    A comprehensive search strategy was applied to PsycINFO and adapted for MEDLINE, EMBASE, ERIC, Education abstracts, CENTRAL, and CINAHL.

    “We used a comprehensive search strategy in order to search an extensive range of relevant electronic databases systematically.The search was applied to PsycINFO, and then adapted for each of the following databases: MEDLINE, EMBASE, ERIC, Education abstracts (H.W. Wilson), Cochrane Central Register of Controlled Trials (CENTRAL), and CINAHL.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    A sensitive literature search was used to avoid overlooking potentially important studies.

    “Since initial background reading suggested that there were not many relevant RCTs, we opted for a sensitive literature search in order not to overlook any potentially important studies (Petticrew & Roberts, 2006).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Petticrew & Roberts, 2006 source

  • May combine two claims.

    Study design and conductreasoning

    The present study used robust variance estimation, which permitted inclusion of all relevant effect sizes in each meta-analysis.

    “Finally, the present study used robust variance estimation, a relatively novel application for meta-analysing data, which permitted the inclusion of all relevant effect sizes in each meta-analysis.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The search strategy used multiple sensitive searches reflecting different elements of the primary question.

    “The search strategy made use of multiple sensitive searches that reflect the different elements of the primary question.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductcitation

    The Cochrane Handbook risk-of-bias tool was applied to all included studies.

    “The Cochrane Handbook (Higgins & Green, 2011) risk-of-bias tool was applied to all included studies.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Higgins & Green, 2011

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    The authors pre-specified five subgroup meta-regression analyses.

    “Five subgroup meta-regression analyses were pre-specified.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    A protocol was submitted to the PROSPERO registry with child language as the primary outcome.

    “A protocol was submitted to the PROSPERO registry for systematic reviews (CRD number: 42017057258) with child language as the primary outcome.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    A tailored data extraction form based on Cochrane Handbook recommendations was developed.

    “A tailored data extraction form (based on the Cochrane Handbook recommendations) was developed in order to obtain all the necessary information from the included studies.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductcitation

    The analysis did not ignore appropriate outcomes, did not make arbitrary decisions about the most appropriate outcome, and did not create synthetic average effect sizes within studies.

    “This approach meant that we did not ignore any appropriate outcomes, or make arbitrary decisions about the 'most appropriate' outcome that matched our criteria, or create synthetic average effect sizes within single studies (Tanner-Smith & Tipton, 2014).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Tanner-Smith & Tipton, 2014 source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    No publication date or publication status restrictions were imposed.

    “No publication date, or publication status restrictions were imposed.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    It has been a decade since the last review of book-sharing interventions on language development was published, and several randomized trials have been conducted since then.

    “It has been a decade since the last review of the impact of book-sharing interventions on language development was published (Mol et al., 2008), and several randomized trials have been conducted since then.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Mol et al., 2008 source

  • Classified as probably not a substantial claim.

    Premisecitation

    Previous reports indicate parental education may affect whether the intervention is delivered with fidelity.

    “Second, studies were categorised according to parental education level because of previous reports that parental education may bear on whether the intervention is delivered to the child with fidelity (Mol et al., 2008).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Mol et al., 2008 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Five pre-specified subgroup analyses on primary outcomes were planned, conditional on sufficient studies for meta-regression.

    “We planned to conduct five pre-specified subgroup analyses on primary outcomes, provided that there were sufficient studies available for a meta-regression analysis.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The effects in Bus et al. (1995) were independent of family socioeconomic status.

    “These effects were independent of family socioeconomic status.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Bus et al., 1995 source

  • Classified as probably not a substantial claim.

    Premisecitation

    Starting intervention earlier in childhood is linked to larger child outcome gains than starting later.

    “For child age, based on indications that intervening early in a child's life is associated with greater improvement in child outcome than later intervention (Mol et al., 2008), child age was dichotomized at the most common point of preschool attendance: those with mean child age at base-line of less than 36 months (younger) and those with mean child age of 36 months or greater (older).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Mol et al., 2008 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Expressive language, receptive language, and caregiver book-sharing competence were analysed separately.

    “Expressive language, receptive language and caregiver book-sharing competence were all analysed separately.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Disagreements over study eligibility were resolved through discussion with a third reviewer.

    “Any disagreements between them over the eligibility of particular studies were resolved through discussion with a third reviewer (PC).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Bus et al. (1995), a meta-analysis of 29 studies, found an overall medium effect size of d = 0.59 on combined language and literacy outcomes.

    “This revealed an overall medium effect size of d=0.59 on the collective language and literacy outcomes (Bus et al., 1995).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Bus et al., 1995 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The Cochrane Risk of Bias Tool was used to evaluate risk of bias within included studies.

    “The Cochrane Risk of Bias Tool was used to evaluate the risk of bias within the included studies.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    PRISMA reporting guidelines were followed.

    “PRISMA guidelines of reporting were followed.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Two reviewers assessed each study for risk of bias, and discrepancies were resolved through discussion with a third reviewer.

    “Two reviewers assessed each study for risk of bias (ND and LH) and any discrepancies were resolved through discussion involving a third reviewer (PC).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    It is surprising that so few evaluations of book-sharing have been carried out in LMICs given that book-sharing programs are inexpensive and easily delivered relative to other ECD interventions.

    “Given that, relative to other more comprehensive ECD interventions, book-sharing programs are inexpensive and can be easily delivered, it is surprising that so few evaluations of booksharing have been carried out in LMICs.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Book-sharing programs cost little and are easy to deliver compared with other early childhood development interventions.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    One explanation is that group delivery provides caregivers with greater social support.

    “A possible explanation is that the group format offers more social support to participating caregivers.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    In low- and middle-income countries, interactive book-sharing could produce substantial gains in early language and possibly broader development.

    “In these settings, interactive book-sharing interventions have the potential to bring about marked, positive effects on children's early language development and, possibly, their wider developmental progress.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    The group-based modality findings should be interpreted with caution because few studies used one-on-one delivery and two of the five individual-delivery studies were very low intensity.

    “However, these findings should be interpreted with caution, given the limited number studies that used a one-on-one format and the fact that that two of these five studies had very low intensity interventions (Goldfeld et al., 2012;High et al., 2000).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Goldfeld et al., 2012 source
    • cites High et al., 2000 source

  • May combine two claims.

    Interpretationreasoning

    Book-sharing programs for caregivers may work regardless of parental education level, including families without an existing book-sharing routine.

    “This is especially important as it suggests that caregiver directed book-sharing interventions can be successful across the full range of parental education, even when there has not previously been a routine of book-sharing (as is the case in families with less educated parents (Bus et al., 1995;Mol et al., 2008)).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Bus et al., 1995 source
    • cites Mol et al., 2008 source

  • Classified as probably not a substantial claim.

    Study design and conductown data

    The study also investigated the effect of these interventions on caregiver book-sharing behaviors, beyond child language.

    “We also went beyond investigating the effect of these interventions on child language by also investigating their effect on caregiver book-sharing behaviors.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationcitation

    Expressive language may have improved more than receptive because book-sharing encourages children to speak actively.

    “The better outcome for expressive language compared to receptive language could be a function of the fact that a central feature of book-sharing interventions is encouraging the child to take an active verbal role in the activity (Mol et al., 2008).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mol et al., 2008 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The search covered educational, psychological, medical, and public-health databases broadly.

    “It involved a wide-ranging search of educational, psychological, medical, and public-health databases.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Both modality models had degrees of freedom below the recommended threshold, so caution is needed.

    “These findings should, however, be interpreted with caution as the degrees of freedom for both models were below the advised level.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductcitation

    The analysis used robust variance estimation, which incorporated several effect sizes per study and outcome and yielded more precise estimates than pooling one effect per outcome.

    “We used a novel statistical approach, namely, robust variance estimation.This allowed for multiple effect sizes per study per outcome to be incorporated, thus providing a more precise estimation than previously achieved using the pooled effect of each outcome (Tanner-Smith & Tipton, 2014).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Tanner-Smith & Tipton, 2014 source

  • May combine two claims.

    Interpretationreasoning

    Dialogic methods often prompt children to speak, which may explain the consistent expressive language advantage.

    “Many of the dialogic techniques employed seek to elicit a verbal response and this promotion of child verbal participation may account for the consistent benefit in child expressive language.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    The categorization into low and high dosage (<60 minutes vs >60 minutes of collective intervention time) was crude but instructive.

    “The categorization into low and high dosage was necessarily crude (<60 minutes vs > 60 minutes of collective intervention time) but it was, nevertheless, instructive”
    Quote in the paperNot a substantial claim
86 of 86

  • The paper relies on this without support.

    No prior systematic review or meta-analysis has focused exclusively on RCTs of interactive book-sharing interventions delivered to carers.

    “There has been no systematic review or meta-analysis that has focused exclusively on the findings of RCTs of the efficacy of interactive book-sharing interventions with carers of children.”

  • The paper relies on this without support.

    Previous reviews focused only on psychological and educational evidence and did not search medical and public health databases such as Medline and EMBASE.

    “Further, previous reviews have focussed exclusively on psychological and educational evidence and not searched medical and public health databases, such as Medline and EMBASE.”

  • The paper relies on this without support.

    Most research on book-sharing interventions has been conducted in the U.S. and other high-income countries, with very little in low- and middle-income countries.

    “Most research and evaluation of book-sharing interventions has taken place in the U.S. and other high-income countries.There has been very little research on book-sharing conducted in low and middle income countries (LMICs).”

  • The paper relies on this without support.

    Intervention dose was expected to vary across studies.

    “The intervention dose was expected to vary between studies, and both brief interventions and more intensive interventions involving multiple sessions were eligible for inclusion.”

  • The paper relies on this without support.

    Book-sharing programs cost little and are easy to deliver compared with other early childhood development interventions.

    “Given that, relative to other more comprehensive ECD interventions, book-sharing programs are inexpensive and can be easily delivered, it is surprising that so few evaluations of booksharing have been carried out in LMICs.”

  • The paper relies on this without support.

    The language and pre-literacy benefits of interactive book-sharing only become fully apparent once children start school and receive formal literacy instruction.

    “The language and pre-literacy skills that interactive book-sharing interventions help to build are only fully realized when a child enters the schooling system and begins to engage with formal literacy instruction.”

  • The paper relies on this without support.

    Random sequence generation and allocation concealment are essential features of randomized controlled trials.

    “These two aspects of study execution are central to the RCT design”
7 of 7

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The lack of statistically significant subgroup results by child age is consistent with prior meta-analyses that showed no age effects among parenting interventions focused during early childhood or spanning broader age ranges.

    “The lack of statistically significant subgroup results by child age are consistent with the few prior meta-analyses that have similarly shown no age effects among parenting interventions focused during early childhood [9,22] or spanning broader age ranges from early to middle childhood [49] or birth to age 18 [62].”

    From Parenting interventions to promote early child development in the first three years of life: A global systematic review and meta-analysis (2021) · cites this paper as [9]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Discussion calls a nonsignificant dosage comparison significant

    Major

    The discussion says the more-intensive interventions produced significantly greater expressive-language benefit, but the reported comparison has p=0.07 and a confidence interval that crosses zero.

  • Sim intervention is classified as intensive despite 15 minutes of training

    Minor

    The methods define medium-to-high intensity as more than 60 minutes of contact time, whereas Table 1 places Sim in that category and lists one 15-minute session.

  • Arnold study total does not match its group counts

    Minor

    The intervention and control counts shown for Arnold sum to 51, while the row reports a total of 41.

    Reported 41 · recomputed 51

  • An included study falls below the stated minimum child age

    Minor

    The eligibility criteria set the child-age range at 12-72 months, but the reported starting-age range for included studies begins at 8.85 months.

  • Video and phone intervention counts disagree

    Minor

    The intervention description counts three video interventions and one phone intervention, but the modality analysis says only three video-or-phone interventions were excluded in total.

5 of 5

Cites

Flagged first, then by citations

Research works printed but not linked: 6

  1. Dunn, 1981, Manual for the peabody picture vocabulary test-revised
  2. Fisher, 1992, The hundred pictures naming test
  3. Higgins, 2011, Cochrane handbook for systematic reviews of interventions (Vol. 4)
  4. Weisleder, 2018, Reading Aloud and Child Development: A Cluster-Randomized Trial in Brazil, Pediatrics, 141, e20170723, 10.1542/peds.2017-0723 · DOI · Paper page
  5. Whitehurst, 2002, Read together, talk together [Motion picture]
  6. World Bank, 2018, World Bank Country Classification by Income

Cited by

Sorted by citations

Related

Authors
Dowdall, N., Melendez-Torres, G. J., Murray, L., Gardner, F., Hartford, L., & Cooper, P. J.
Affiliations
  • University of Oxford (GB)
  • Cardiff University
  • University of Reading (GB)
  • University of Cape Town (ZA)
Corresponding author
Dowdall, Nicholas
ORCID
Venue
Child Development
Publisher
Oxford University Press (OUP)
Published
2020-03-01
Volume
91
Issue
2
Pages
e383-e399
ISSN
  • 0009-3920 (ISSN-L)
  • 1467-8624
DOI
10.1111/cdev.13225
OpenAlex
W2908423519
Cited by
391
References
62
Citations by year
2026: 89 · 2025: 74 · 2024: 81 · 2023: 58 · 2022: 35 · 2021: 33 · 2020: 18 · 2019: 2 · 2018: 1
FWCI
4.67
Type
journal article
Language
en
Open access
Licence
Locations
  • Child Development (publishedVersion)
  • PubMed (publishedVersion)
  • CentAUR (University of Reading) (acceptedVersion)
Funders
  • Wellcome Trust
  • Cancer Research UK
  • British Heart Foundation
Topics
  • Social Sciences › Psychology › Developmental and Educational Psychology › Language Development and Disorders
  • Social Sciences › Psychology › Developmental and Educational Psychology › Reading and Literacy Development
  • Social Sciences › Education › Early Childhood Education and Development
Keywords
shared book reading, picture books, parent-child interaction, child language acquisition, expressive language, receptive language, intervention dosage, early childhood
MeSH terms
Books, Child, Child, Preschool, Female, Humans, Interpersonal Relations, Language Development, Male, Reading, Vocabulary

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read under definition version 3 (extract-v3) on 2026-09-28.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-09-28.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review